Professional Documents
Culture Documents
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
Cambridge Center for Behavioral Studies (CCBS) is collaborating with JSTOR to digitize, preserve and extend
access to Behavior and Philosophy.
http://www.jstor.org
Behavior
and Philosophy,
2006 Cambridge
A Review
of Max
Hacker's
Philosophical
R. Bennett
Studies
to Science?
is Philosophy
Of what Value
and P. M.
Foundations
S.
of
Neuroscience
University
Jose E. Burgos
CEIC
of Guadalajara
JohnW. Donahoe
University ofMassachusetts,
Amherst
jdonahoe@psych.umass.edu.
71
This progress has given its practitioners, some of Nobel Prize fame, a high sense of
often expressed
as a celebration
self-confidence
of its independence
from
In
celebration
into
derision.
this
book,
philosophy. Alas,
degenerated
mocking
strike back?and
with a vengeance?convincingly
however, philosophers
showing
that neuroscience
is not as healthy as it seems. One of itsmost appealing areas,
philosophy
psychology.
in particular
that can serve as a guide for other disciplines,
not
driven
unreserved
is
of
course,
recommendation,
approval
by
of science
Our
(see later)but the importanceof the topic and the scholarlymanner inwhich
the soul. Chapter 2 focuses on the work of Sherrington and his disciples (Adrian,
Eccles, and Penfield), whom the authors regard as Cartesian dualists. In Chapter 3,
the authors sketch their main criticism. Current cognitive neuroscientists are not
substance dualists but repeatedly commit Descartes' mistake, what the authors call
"meros"
fallacy." The term refers to mereology (from the Greek
with
deals
or
that
of
the
branch
part-whole
ontology
"portion"),
meaning
"part"
see Simons,
relations (for the definitive technical treatise on mereology
1987).
to this criticism and an
This chapter also includes rebuttals of some objections
the authors use to
outline of Wittgenstein's
argument, which
private-language
terms are learned.
propound a view of how themeanings of ordinary psychological
the "mereological
analysis"
(p. 378),
a method
(Chapter
72
examine
Dennett,
Chalmers,
critically the views of philosophers of mind such as McGinn,
more
on consciousness.
Part IV waxes
and Searle
philosophical,
and
philosophical
Daniel Dennett
neuroscience
and are dedicated
and John Searle.
are also
two appendices
14). The
(Chapter
to criticisms of the methodological
of
proposals
neuroscience":
doctrine of Nemesius
influence on cognitive-neuroscientific
research for the next
philosophical
three centuries to the present time. Although current cognitive neuroscientists have
substance dualism, they keep ascribing psychological
attributes
largely abandoned
themain
to parts of human
division
73
The mereological
fallacy can be seen as a sort of compound of these two
restricted to ordinary psychological
terms and concepts. In cognitive
the
is
manifested
in expressions where
neuroscience,
fallacy
paradigmatically
are asserted to perceive, believe, know, reason,
brains and brain hemispheres
imagine, remember, feel, and be aware or conscious. The problem with these
fallacies,
expressions
nothing,
so
mereological
through more
is not that they are false but that they are meaningless.
They assert
are
not
are
assertions.
nonsensical
they
They
gibberish. The
or
a
error
not
can
is
factual
theoretical
that
be corrected
fallacy, then,
For example,
consider
the terms "sight" and "belief," with all their
in Aristotelian
variants. As
grammatical
epitomized
hylomorphism,
they are
see
a
to
to
creatures.
"/
used
refer
whole
One
red
ordinarily
light,"
ordinarily says
not "my brain sees a red light," and "you believe
it is raining," not "your brain
it is raining," and so on. The two senses are conceptually connected (e.g.,
"I believe I saw a red light"). One might want to use "sight" more precisely, to
refer to a part of a creature, like its striate cortex. In the interest of clarity, the new
believes
sense should be introduced through an explicit definition. For instance, one should
at the outset thatfor thepurposes of the analysis, "sight" will be defined as
"a temporary activation of one or more neurons in striate cortex correlated with a
radiation of a certain wavelength."
temporary presence of an electromagnetic
declare
Under
this definition,
expressions
like "My
brain
saw
red
light" become
meaningful.
The authors do not prohibit this kind of redefinitional move per se. They have
or quasi-technical
no problem
of ordinary
redefinitions
with
technical
a
common
are
terms.
scientific
Such
redefinitions
practice that the
psychological
authors do not dispute:
There is nothing unusual, let alone amiss, in scientists introducing a new way of
talking under the pressure of a new theory. If this is confusing to the benighted
readers, the confusion can easily be resolved. Of course, brains do not literally
think, believe, infer, interpret or hypothesize, they think*, believe*, infer*,
interpret* or hypothesize*. They do not have or construct symbolic
representations, but symbolic representations*, (p. 74)
Our criticisms of the mereological
fallacy in neuroscience do not preclude
neuroscientists from using the verbs 'to think', 'to believe', 'to perceive', 'to
remember' in new ways according to conditions other than the received
conditions of theiruse, as long as they can explain what these new uses mean.
They can, if they so wish, redefine 'thinking', 'believing', 'perceiving',
'remembering', and give a meaning to the phrases 'My brain thought that itwas
better to keep silent', 'Your brain believes that it is Tuesday tomorrow', 'His
brain perceived that she was smiling', or 'Her brain remembered to go home.'
(p. 384)
74
these answers
either, although
redefinitions
formation rules would have to be stipulated, the conditions for the correct
application for these innovative phrases would need to be specified, and the
logical consequences of theirapplication would have to be spelled out. (p. 384)
New
The result of this kind of task, however, would be a system of concepts quite
ones that motivated
different from the ordinary psychological
the analysis in the
firstplace.
The authors' criticism is that cognitive neuroscientists neither have done, nor
seem to want to do, the additional work. Rather, "they are trying to discover the
for thinking, believing, perceiving
neural basis
and remembering?not
for
something else" (p. 384). This
dubious is to try to accomplish
is
task, of course, is perfectly legitimate. What
as
it by construing brains and brain hemispheres
thinking, believing, perceiving, and remembering. Such constructions do not allow
us to understand human behavior any better than construing clock cogs as giving
the time allows
understanding.
descriptions of experimental findings.
In our example, the authors would
connections
imposed by "sight*" relative to those imposed by the conceptual
between "sight" and "belief." The meaning of "sight*" is thus mixed up with that
of "sight." It is this kind of semantic muddle
that the authors regard as incoherent
and confused, and it is what
ever since Nemesius.
a paradigmatic
to
fallacy in cognitive neuroscience:
example of the mereological
terms
the
assert about a human brain hemisphere what, under the meanings
of
used, makes sense only to assert about whole humans. The problem, again, is that
75
such
expressions
are meaningless,
for
"remembering," "reasoning,"
technically redefined, let alone their new conceptual
the only meanings
they can have when
Consequently,
"thinking,"
"perceiving,"
have not been
interrelationships specified.
used are the ordinary ones.
the
human beings. Hence,
these meanings
apply only to whole
expressions assert nothing and they do not contribute to our understanding
research at all. They only add confusion.
However,
Instead of Sperry's
account,
of this
the following:
What
description
radiations
cortex of each hemisphere, how they are functionally related, and how their
normal relations are disrupted by commissurotomy. This explanation can be made
visual
of thedifferent
as refinedas wished by appealing to the structure
and functioning
that constitute
microcircuits
molecular
levels. This
phenomenon,
let alone
each
possibility,
the normal
76
muddle
is interwoven
neuroscience.
more accurate
invalidate
nor
neuroscience,
can
a term
not
research on themind-body
Our
nexus.
Concerns
basic
unclear whether
souls being parts of humans, or humans being composites of souls and bodies.
Second, there is an odd tension between the authors' Wittgensteinian
stance, which
is characteristically
and
their
embracement
of
Aristotelian
anti-essentialistic,
which
is characteristically
hylomorphism,
case against
unconvinced
the
authors'
by
concern
turn.
elaborate each
in
Cartesian
The
Dualism
and theMereological
essentialistic.
explanatory
were
Third, we
Let us
reductionism.
Fallacy
authors'
neuroscientists,
charge of "crypto-Cartesianism"
against current cognitive
most
the
of
the
book
perhaps
(see pp. 111-114, 233
striking aspect
77
requires
further examination.
the human being as a whole. The fallacy crept into cognitive neuroscience
through
on Sherrington and his disciples, who also ascribed
the influence of Descartes
attributes to themind. Substance dualism was eventually abandoned
psychological
in favor of materialism,
but parts of human beings, typically brains and brain
as the bearers of
came
to replace
and remained
the mind
hemispheres,
attributes.
psychological
Our concern here is whether the mereological
fallacy can be committed
within
it is equally unclear
Consider
for
has acts, a symphony has movements, and so on. Events thus exist incompletely in
in time during their occurrence. A baseball game in its inning four
any moment
exists incompletely during that inning, for some of its parts (innings one, two, and
A basic intuition about parts is that they are smaller than thewholes of which
they are parts. An arm is smaller than a body, an atom smaller than a molecule.
It is also
honored
of a mereological
event is a shorter event. An inning is shorter than a baseball game; an act is shorter
a soul cannot be intelligibly said to be shorter than a
than an opera. Obviously,
human in Cartesian dualism, for in this doctrine souls are immortal, bodies are
mortal, and humans are interactions between souls and bodies. Souls, then, cannot
be
78
Cartesian
done without
Cartesian dualist need not go down this slippery slope. Cartesian dualism would
not seem to depend critically on mereological
talk of souls being parts of humans
or humans being composites
fusions, mixtures, or unions) of
(or combinations,
seem to be entirely dispensable
in Cartesian
souls and bodies. Such talk would
material
relation. Descartes'
talk of souls as parts of humans and humans as
mereological
or
unions
souls
of
and
bodies can thus be safely dismissed as careless
composites
and ontologically
Such talk represents no significant aspect of
inconsequential.
Cartesian
dualism.
These
dualism.
causally
Cartesian
the mereological
in Cartesian dualism.
fallacy is entirely dispensable
To be sure, the fallacy can be committed in current cognitive
for the
neuroscience,
position. On
considerations
79
in current
in a better
intellectual
faultless genius, a myth that confuses brightness with perfection. A moral of the
in this respect is that even geniuses can and do make blatant mistakes.
theNobel Prize, the archetype of the scientific genius, does not mean that
Winning
thewinner is right in everything he or she says, writes, or does. The Nobel Prize is
book
for an outstanding
recognition
intellectual flawlessness.
not
in science,
achievement
certificate
of
In itsmost
Essentialism
second
commitment
versus Wittgensteinian
concern
arises
toWittgenstein's
from an
odd
Anti-Essentialism
tension
between
the authors'
(as expounded
in his
basis
of the essential,
defining
functions of a
essentialistic
the assertions
that "[psychological
predicates
are predicates
that
apply essentially to thewhole livinganimal, not to itsparts" (p. 72) and "[the
ability
and be aware
of them] is essentially
dependent
upon
80
as
"Essence
is expressed
wrote:
by grammar"
(Philosophical
Investigations,
?371),
he also
Consider
the authors'
talk of conceptual
commonality
Concepts are abstractions from the use of words. The concept of a cat iswhat is
common to the use of 'cat', 'chat', 'Katze', etc. (p. 65, our italics)
The words 'cat', 'chat', and 'Katze' are symbols in three different languages, all
ofwhich express one and the same concept, (p. 345, our italics)
This
to admit an account
in terms of commonality
and sameness. The
changeable
to
sense
authors also make reference
the logicians'
of "qualitative
identity" (p.
it. This reference too is puzzling when
96n), apparently without disowning
to Wittgenstein's
turn from formal logic in the Tractatus
to
compared
(1961)
ordinary
language
Wittgenstein
linguistic, not ontological. Universals may well exist (although demonstrating their
existence
is no trivial matter); however, for better or worse,
them
expressing
linguistically requires a highly formalized language that extremely simplifies and,
to this extent, departs considerably
from ordinary language,
the focus of the
authors' analysis.
Explanatory
Reductionism
we
were
by the authors'
unimpressed
rejection of explanatory
This rejection is largely independent of the authors'
can thus
of conceptual
in cognitive neuroscience.
confusion
One
diagnosis
with
the
and
still
with
the
coherently agree
diagnosis
disagree
rejection. Avoiding
themereological
fallacy does not commit oneself to explanatory anti-reductionism,
nor does explanatory anti-reductionism entail the fallacy.
Thirdly,
reductionism
(pp. 355-366).
rejection in question has two aspects. On the one hand, the authors claim
that human action cannot be explained in terms of neural laws because there are no
The
psychological
81
exactly what
when
they deny
that "there is
historical
explanations the authors mention are good candidates for probabilistic laws.3
on the other hand, even if reasoned human action were
But no matter?for,
to
neural laws, the resulting explanations would be inferior
reducible
explanatorily
to those that appeal to the behaving person's reasons:
call on Jack only to find him out.We ask where he is, and are told he has
gone to town.We want to know why, and are told that it is his wife's birthday,
We
82
the first and third questions with a resounding "yes." The authors'
at social
of Jack's behavior
arbitrarily stops the explanation
and conventions. But surely many (us included) are further puzzled by
answer
answer
negative
practices
the practices
and
conventions
themselves.
can thus be
question
are
those practices and conventions
The
second
in terms of explanations
of how
are
how
instantiated in specific individuals, what the
they
acquired
of
the
similarities
and
differences
observed among them are, and so on.
origins
to these questions will certainly deepen our understanding of reasoned
Answers
answered
and maintained,
(sans
themereological
fallacy) may
have much
on what
the authors mean
and
course, much depends
by "deepen"
can
Here
the
authors'
be
turned
toward
Wittgensteinian
"understanding."
approach
Of
of "deepen"
is "to extend well
inward from an outer
Online
(Merriam-Webster
By referring to certain inward
Dictionary).
and their functioning (brain hemispheres,
anatomical macro- and microstructures
them. One
surface"
ordinary use
to achieve a
in thought as though expressed, to have the power of comprehension,
or infer
or
to
of
believe
of
the nature, significance,
grasp
explanation
something,
or
a
tolerant attitude toward
sympathetic
something to be the case, to show
83
Again, much depends on what the authors mean by "clearer" here. Ordinarily, to
render clearer is to free from opaqueness,
ambiguity, or indistinctness, to make
or
render human behavior
unclouded.
Neuroscientific
transparent
explanations
as
more
to
is publicly
what
lies beyond what
transparent insofar
they refer
or
in ordinary situations. They will also make
it less ambiguous
observable
human
behavior
may well
be due
will
substrates.
Granted, using water effectively in everyday life (for quenching one's thirst,
bathing, boiling food, making ice, dissolving, etc.) does not require knowledge of
intuitive knowledge of the reactions of water suffices for
quantum mechanics?our
that?but
itwould be grotesque to regard quantum-mechanical
explanations of the
reactions of water as somehow
the latter
inferior to intuitive ones just because
are
suffice for everyday life. There
non-ordinary (scientific and technological) uses
that require quantum mechanics.
They admittedly are far removed from
common
life
and
but
this
does not make quantum-mechanical
sense,
everyday
ones.
to
At
inferior
intuitive
worst, they are inferior relative to
explanations
uses
are more legitimate or important
to
who
is
but
that
such
say
ordinary uses,
of water
with
reasoned
to intuitive explanations.
in everyday
It surely seems
human action
by cognitive neuroscience,
not know any better. When
one can rightly ask: "Who needs cognitive neuroscience
to understand
this?"
for that. However,
this
Indeed, we need not bother with cognitive neuroscience
does not logically exclude the possibility of future circumstances where knowledge
of such laws will be required. Nor
is this possibility
logically guaranteed, of
course, and herein lies the predicament. Our present ignorance prevents us from
making any certain predictions in this respect, one way or the other. It is not even
certain thatwe will be able to discover the relevant psychological
laws, or that they
to
out
to
is
exist. The only way
try,with all the concomitant risks (waste of
find
time, money, and energy), but this is the way science works.
involves a great deal of trial and error (and success as well).
Science
84
inevitably
Review
of Bennett
& Hacker
Concluding Remarks
What
should
with
(p. 142)
Neural
implicated
in the exercise
of the relevant
and
between philosophy
there is the authors'
Second,
sharp disanalogy
the
science. In particular, they claim that philosophy
philosophy of
(specifically,
across
the boundaries
between
which
neuroscience,
"operates
cognitive
. .
to the a priori,
(p. 2). This claim appeals
.neurophysiology and psychology"
versus
the
of ordinary
character of the philosophy
language
non-empirical
the
authors
that
of
neuroscience.
On
claim
character
this
basis,
cognitive
empirical
philosophy does not, cannot, and should not suggest new experimental research in
405). More
elaborately:
85
much
of
rationalism,
innatism,
and
(more modernly)
nativism,
all
of which
nexus. The
posteriori distinction is only half the story of the philosophy-science
can of
other half is the analytic-synthetic distinction, a veritable philosophical
worms that the authors open and close very quickly (p. 438n). This distinction is
in philosophy
truth" is a standard alternative
"conceptual
important because
expression for "analytic truth." In this use, analytic truth is a matter of linguistic
convention. A statement or proposition is analytic if and only if it is true entirely in
virtue of the meanings
of the terms that constitute it (e.g., "all bachelors are
"all vixens
British empiricists nor the logical positivists had any problem whatsoever with it).
Behavior
scientists who take experience as themain or only source of knowledge
may thus embrace the a priori without incurring any intellectual liability. Quite the
contrary, they could benefit greatly from the kind of analysis presented in the
insofar as conceptual
to science as experimental
clarity is as essential
rigor. It is unclear whether or not radical behaviorists would be convinced by this
line of argument. They might as well agree with Quine (another philosopher often
book,
86
to have
distinction
altogether.
Skinnerian
ensue.
much
richer thanWittgensteinian
analysis
87